Lovable is not growing because it found one clever channel.

It is growing because it built a system where:

• the product creates delight

• delight creates sharing

• sharing creates trust

• trust creates growth

In Elena Verna’s framing, growth is a trust problem now.

When software becomes easier to build, functionality stops being enough. People can copy features. They can build alternatives. What they cannot copy as easily is:

• trust

• brand

• speed

• relevance

• emotion

• community energy

• product-led word of mouth

That is the real growth engine.

We broke down Lovable’s growth into a system that can be understood and, in parts, applied whether you’re early or already scaling.

1. Why growth is becoming a trust problem (not a channel problem)

2. How to use your product as the main growth channel

3. Why you need a “minimum lovable product,” not just a functional one

4. How to use founder-led social to drive early traction

5. How to turn employees into a distribution engine

6. Why AI-native companies blur roles (and why it matters)

7. Why avoiding the biggest mistake in early-stage growth (paid too early)

8. How to sequence organic and paid growth the right way

9. Why most communities fail (and how to build one that works)

10. How to turn users into your most effective growth channel (UGC)

11. How to define activation around real value, not vanity metrics

12. How to use continuous launches to drive retention and relevance

13. How to stay top-of-mind in fast-moving markets

14. How to design monetization for bursty AI usage

15. Why brand becomes critical as products commoditize

16. How to use brand and offline campaigns effectively (and when)

17. Why leveraging creators as a growth channel

18. How to turn free product moments into growth campaigns

19. Why speed acts as a trust signal

20. How to compete with incumbents using speed, culture, and distribution

Why it’s valuable

This guide connects a lot of scattered ideas into a more coherent model.

You can use this to:

• think about product as distribution, not just utility

• treat free usage as awareness, not just lost revenues

• prioritize organic pull before scaling paid

• design for sharing and visibility, not just usage

• avoid scaling a system that hasn’t earned trust yet

The full breakdown is available for paid subscribers. If you’re not one yet, you can join here:

Subscribe now

Read more

How is it even remotely normal that a 19-year-old scaled an app to ~$50M ARR in 18 months…

…then sold 100% of it to the category leader…

…with zero VC funding???

No seed. No Series A. All bootstrapped.

Just raw distribution and the most ruthless execution I’ve ever seen from any founder his age.

And now he just dropped an interview where he lays out exactly how they did it.

The actual mechanics: Attribution hacks, creative frameworks, campaign structures, UGC systems…

This is not an “AI startup” story.

This is a masterclass in modern consumer growth 🔥

Cal AI is a consumer app that lets you take a photo of your food and instantly get calories and macros.

Just:

📸 Snap → 📊 Calories + macros → Done.

That’s it.

In one of the most competitive consumer categories on earth (calorie tracking) this 19-year-old founder built this to:

• $30M revenue in 2025

• $5.7M revenue in January alone

• ~$50M ARR / run-rate

• Zero funding

• Then sold 100% of it to MyFitnessPal

Eighteen months. Here’s how:

What We’ll Break Down

1. Start With a Product That Demonstrates Itself in 2 Seconds

Why Cal AI built a product that demonstrates itself instantly on video.

2. Distribution Before Product

How they validated a content ecosystem before writing a single line of code.

3. The Brutal Early Days: 500 DMs → 2 Creators

How manual influencer outreach became their first growth engine.

4. The Economics That Made Influencer Marketing Profitable

The RPM vs CPM framework that made creator marketing profitable.

5. The Negotiation Tactics To Lower Creator Costs

How bundling, view guarantees, and DMs cut influencer costs dramatically.

6. The UGC Creative Strategy & How to Copy Them (INSANE🔥)

Why native content beats ads — and how Cal AI engineered viral creator videos.

7. Turning the Comment Section Into a Funnel

How pinned comments and fast replies turned curiosity into installs.

8. Building a Creator Discovery Machine

How they trained TikTok’s algorithm to constantly surface new influencers.

9. Scaling to 150 Creators

Why retainers created predictable distribution and lower acquisition costs.

10. The Paid Ads Multiplier: $2M/year to $50M/year

How influencer videos became the creative engine for profitable Meta and TikTok ads.

11. Using Apple Search Ads to Attack Competitors

Using Apple Search Ads to intercept competitor demand.

12. Understand & Optimize Conversion Engineering

How onboarding psychology and trial design increased subscription conversion.

Apply to The Pitch by Deel now

A teenager built a consumer subscription app to ~$50M run-rate in a year and a half, bootstrapped, in one of the most competitive categories on earth.

And then sold it to his biggest competitor. explain here what it is

Let’s break down exactly what he revealed, and why this might be one of the most instructive consumer growth case studies of the last few years 🔥

How Cal AI Went From a High School Project to ~$50M ARR 🔥

The Exact Growth Systems They Used (In Order)

Read more

Everyone is arguing about how powerful AI is getting. Wrong conversation!

What matters now is where leverage survives once intelligence is cheap, and where it doesn’t.

This week, in OpenAI’s Town Hall, Sam Altman wasn’t making announcements, but he was outlining constraints. The kind that quietly determine which startups work and which ones never compound.

Below are 12 moments from that conversation that actually matter if you’re building: reordered, compressed, and interpreted for people who care about product, distribution, and durability.

First impressions matter.

With Framer, early-stage founders can launch a beautiful, production-ready site in hours. No dev team, no hassle.

Join hundreds of YC-backed startups who launched here and never looked back.

Pre-seed and seed-stage startups new to Framer can enjoy:

• One year free: Save $360 with a full year of Framer Pro, free for early-stage startups.

• No code, no delays: Launch a polished site in hours, not weeks, without hiring developers.

• Built to grow: Scale your site from MVP to full product with CMS, analytics, and AI localization.

• Join YC-backed founders: Hundreds of top startups are already building on Framer.

Apply to claim your free year →

What We’ll Break Down 🔥

1. Jevons Paradox for AI: Cheaper Intelligence = More Software

2. WhyGTM Is Still the Bottleneck

3. When the Platform Doesn’t Know the Interface
Why OpenAI’s uncertainty signals category creation, not stability.

4. The 100× Cost Curve Is the Real Disruption
How infrastructure deflation unlocks entire categories before features do.

5. From Tools to Solved Problems
Why continuously generated software changes the unit of value.

6. The Founder Trap: Betting on Future Model Upgrades
Why relying on model progress is not a moat.

7. The Agent Trust Layer: Why Verification > Autonomy

8. Integrations Are Not Defensibility
As models learn tools instantly, advantage shifts to workflows and proprietary data.

9. The Hidden Risk Curve of Agents
Permission creep, rare failures, and the rise of monitoring infrastructure.

10. Memory as the New Switching Cost
Why persistent context, not raw intelligence, compounds defensibility.

11. Org Design in an Abundance Era
Leverage per employee becomes the only metric that matters.

12. Assume 100× Everything
How to think like a builder in a world where intelligence is no longer scarce.

Let’s get into this 🔥

Read more

Most founders don’t fail because the product sucks.

They fail because they keep acting like the best employee when the company needs a CEO.

At scale, “being productive” stops meaning what you can do today. It means what the company can do without you. If you don’t make that switch, you become the bottleneck. hen the board, the market, or burnout makes the decision for you.

This isn’t motivational. It’s statistical. In a study of 1,156 VC-backed U.S. IPOs, 40.8% changed CEOs between the first VC round and the IPO. And HBR notes founder-CEO transitions are 2–3x more likely to fail or cause a performance drop than transitions between nonfounder CEOs.

So this guide is about one thing: killing the “founder mode” habits that stall growth, and replacing them with CEO systems.

It includes:

1. The 4 CEO fundamentals founders learn too late

2. The step-by-step process to remove yourself as the bottleneck

PMF is read by more than +1000 different VC firms & tens of thousands of the fastest growing startup founders including multiple YC founders.

Want to put your company in front of them?

Sponsor PMF

1. The 4 CEO fundamentals founders learn too late

Read more

Naval Ravikant predicted a future where small, specialized startups dominate niche problems.

We’re now seeing that vision come to life with AI Solopreneurs: professionals who use their expertise in tandem with AI to produce massive results: all under a single-person operation.

With Framer, early-stage founders can launch a beautiful, production-ready site in hours. No dev team, no hassle.

Join hundreds of YC-backed startups who launched here and never looked back.

Pre-seed and seed-stage startups new to Framer can enjoy one year free: Save $360 with a full year of Framer Pro, free for early-stage startups.

Apply to claim your free year →

Subscribe now

What’s an AI Solopreneur?

It’s pretty simple: imagine a consultant or freelancer who combines human expertise with AI tools—enabling them to handle the work of an entire team without building a massive organization.

They enjoy:

• Top-tier results for clients at competitive rates

• Manage multiple projects without burnout

• Easier scaling—AI handles tedious tasks so you can focus on what really matters

To productize yourself is to turn your skills, experiences, and insights into repeatable, marketable offerings—like coaching programs, digital courses, or done-for-you services.

By combining this approach with AI-driven automations, you can handle more clients, solve bigger problems, and truly work less while delivering great value.

Three Key Pillars: Marketing, Craft & AI

Despite the tech transformation, the fundamentals remain: you must excel at Marketing and at a specific Craft.

1. Marketing

Step 1: Choose Your Tribe (Your Niche)

Pick a hyper-specific niche that resonates with both your skills and passions. Examples:

• “AI-driven sales funnel optimization for B2B SaaS”

• “AI-powered copywriting for health & wellness brands”

• “Fitness coaching for busy professionals”

When you specialize, you stand out, command higher rates, and attract clients who see real value in your expertise.

This is a great example of the latter. Blogging for devs: a course of blogging for developers. Super niche in:
1. The target
2. The skill it teaches

Step 2: Find Where Your Tribe Hangs Out

• Online Communities: LinkedIn groups, Subreddits, Slack channels, Discord servers

• Social Media: X (Twitter), LinkedIn, Instagram, or TikTok—only pick the ones your ideal clients actually use

• Offline Events: Industry expos, local meetups, conferences—yes, old-school networking still works

Step 3: Create (Market)

2. The Craft

The idea is pretty simple:

Choose the skill/craft you’ll be monetizing and practice it as much as possible to become as good as possible.

If you work on it enough you can become the best of the world in less than 5 years.

Now, up to this point you have done the same any small professional would do to get by, but how can you scale this?

3. Leverage AI to Multiply Your Output x50

Imagine having 50 virtual assistants working around the clock. AI tools can help you:

• Learn how to Generate Content at Scale: Use ChatGPT or other large language models to draft blog posts, email campaigns, or social media updates. Then edit, refine, and humanize that content so it remains authentic to you.

  For example: I created a GPT to build Linkedin Viral Infographics like this👇

• Automate Research: Tools like Bing Chat or specialized web-scraping AIs can sift through data or trends in minutes—what would have taken you hours (or days) now takes moments.

• Streamline Communication: Let AI handle meeting summaries, customer service chatbots, or quick email replies—so you’re free to focus on creative or strategic tasks.

• Optimize Workflows: Link AI with no-code platforms (e.g., Zapier, Make) to handle repetitive tasks (scheduling, file organization, lead capturing), reducing human error and saving you massive time.

The result? You can deliver 50x more value—creating proposals, prototypes, and entire marketing campaigns in a fraction of the time.

While AI churns out drafts, data, or routine follow-ups, you refine high-level strategy and maintain the human touch. This synergy between tech and talent is the secret sauce for standing out in a crowded market.

Access Now 200 Ways to Make Money with AI

Check it out with the free trial (no strings attached), or jump straight into premium if you’re ready for the full scoop👇

Read more

(WARNING🚨: This is a brutally direct and practical execution guide)

Most startup advice on X or LinkedIn fails for one simple reason: it’s written to sound smart and get views, not for getting results.

This is different.

This is a step-by-step playbook for going from idea → first revenue in under 21 days, built by compiling what actually worked early for companies like Reddit, Stripe, Buffer, and Superhuman, alongside the early stage advice of Paul Graham.

It’s designed to be executed by one person: no team, engineers nor audience.

Every step is specific backed by what the founders of great companies did.
Every tool is named.
Every decision has a clear “move on” or “stop” signal.

If you execute this properly, one of two things will happen within 21 days:

1. Someone pays you for what you’re building

2. You get hard proof of why they won’t, before wasting months building the wrong thing

Both of these are incredibly valuable if you want to build long-term.

What you’ll actually get from this guide

Most founders duct-tape an MVP, demo it to friends who won’t say no, call it “validation,” and forget about it in under a week.

This is not that. By the end, you will have:

• a clearly defined buyer (real people, real companies)

• a priced offer you’ve said out loud to prospects

• a list of 50–100 potential customers you personally contacted

• transcripts of real conversations with buyers

• at least one explicit yes or no to paying

This guide is for solo founders, first-time builders, and operators validating an idea who are tired of vague “just ship” advice with no instructions.

First impressions matter.

With Framer, early-stage founders can launch a beautiful, production-ready site in hours. No dev team, no hassle.

Join hundreds of YC-backed startups who launched here and never looked back.

Pre-seed and seed-stage startups new to Framer can enjoy:

• One year free: Save $360 with a full year of Framer Pro, free for early-stage startups.

• No code, no delays: Launch a polished site in hours, not weeks, without hiring developers.

• Built to grow: Scale your site from MVP to full product with CMS, analytics, and AI localization.

• Join YC-backed founders: Hundreds of top startups are already building on Framer.

Apply to claim your free year →

The 21-Day Breakdown:

Days 0–3: Kill or Commit 🔪

How to FIND and validate the problem, define the buyer, and decide if the idea deserves 21 days of your life. Follow Paul Graham’s core advice.

Days 4–5: Define the Offer & Price It 📦

Turn the idea into something concrete people can say yes or no to. Follow Buffers playbook regarding pricing before launching.

Days 6–13: Build the Smallest Sellable Version 🧱

Create only what’s needed to deliver value, NOTHING more. Learn from Reddit’s early “product faking” strategy.

Days 14–16: Get in Front of Real Buyers 💬

How to find and target specific companies, message potential customers, and start conversations organically. Stripe’s playbook: how they talked to users before thinking about sales.

Days 16–21: CLOSE & Deliver 🤝💰

Close, close, close. Ask for money, onboard personally, and extract the truth from outcomes. Follow Superhuman’s playbook: onboard users personally and how they charged from day one

Lets get to it🔥

Days 0–3: Kill or Commit

Day 0: How to Find a Real Problem

Read more

AI is about to split companies into 2 categories:

1) The ones still paying 2020 prices for 2026 output
2) The ones rewriting their cost structure with AI

Because what used to take a $100,000 budget now costs a few hundred bucks.

Not in theory. In practice:

• Design → from “weeks + agency” to “hours + tools”

• Copy → from “copywriter + revisions” to “drafts in minutes”

• Research → from “analyst time” to “instant synthesis”

• Sales upgrades → from “enablement team” to “always-on coaching + personalization”

• Customer support → from “more headcount” to “resolution automation”

Whole teams are getting compressed into software.

So this isn’t “tech trend” content.

This is your P&L getting rewritten.

And the most dangerous part?

Your competitor isn’t waiting for a perfect strategy deck.
They’re learning this right now, quietly, every day:

• Which workflows to automate

• Which roles to augment

• Which costs to delete

• Which outputs to 10x

In 12 months, a lot of markets will look “the same” from the outside…

…but one company will be running with a cost base that makes everyone else uncompetitive.

Hey everybody, welcome to PMF.

In this newsletter you’ll find the tools and ways some of the smartest founders are using to go create content in multiple formats and distribute it massively at a fraction of the cost of what it used to be.

1. AI Video generation tools

2. How AI Influencers generating millions of views

3. How to distribute your content with UGC (how ElevenLabs is doing it)

   

Subscribe now

Join OpenAI for Startups

Sponsor Product Market Fit & get your product in front of thousands of founders worldwide? Reach out to g@guillermoflor.com

Trending right now: The Data Room Template VCs love, 27 Most Promising AI Startup Pitch Decks Backed by Top Investors in 2025

1) AI Video Generation Tools

Read more

Great founders study problems.

Airbnb didn’t start with “let’s build a marketplace for short-term rentals.”
Stripe didn’t start with “let’s build payments infrastructure.”
OpenAI didn’t start with “let’s ship a chatbot.”

They all started with something much more primitive:

“This is broken. And nobody has fixed it properly.”

In this piece, I’ll do two things:

1. Break down the core problems that inspired some of the biggest tech companies in the world (Airbnb, Stripe, Amazon, OpenAI)

2. Share 15 unsolved problems that I believe can still produce decacorns heading into 2026

If you’re a founder, this is how you should be training your brain.

Join OpenAI for Startups

Sponsor Product Market Fit & get your product in front of thousands of founders worldwide? Reach out to g@guillermoflor.com

Trending right now: The Data Room Template VCs love, 27 Most Promising AI Startup Pitch Decks Backed by Top Investors in 2025

Part I – The Problems Behind Iconic Startups

Read more

The X algorithm is a pipeline. A very literal set of steps:

• pick a bunch of tweets

• throw most of them away

• predict what you’ll do on the rest

• score them

• show you the top few

• learn from what you did

We went through the actual system that powers the For You feed and translated it into normal language.

If your grandma can understand how Netflix recommends shows, she can understand this.

In this newsletter you’ll get the full playbook to go viral on X + 50 tweet templates designed for the actions X rewards

1. How X builds a “taste profile” for every user

2. How X pulls tweets from two different buckets

3. How out-of-network discovery really works

4. Why most tweets never even get scored

5. How X predicts what a user will do when they see a tweet

6. Why replies and reposts matter more than likes

7. How negative signals quietly destroy reach

8. Why posting too much can actually hurt your distribution

9. Why the first 15–30 minutes decide a tweet’s fate

10. How the feedback loop compounds winners

11. A simple founder playbook you can actually follow

12. 50 tweet templates designed for the actions X rewards

Let’s get to it 🔥

Join OpenAI for Startups

Sponsor The AI Opportunity & get your product in front of thousands of founders worldwide? Reach out to g@guillermoflor.com

Trending right now: How Base44 was bootstrapped and sold for $80M in 9 months

Get access to this and much more with the PMF subscription

💰 The Ultimate Investor List of Lists (1,000 Family Offices, 1,000 Active Angel Investors, 160+ No-Warm-Intro Investors, 3,000 LPs…)

🖼️ +50 Pitch Decks of World’s Most Sucessful Startups

🖼️ 27 Most Promising AI Startup Pitch Decks Backed by Top Investors in 2025

🔥 How to Get Into YC: The Spring 26’ FULL GUIDE

Get 40% off forever

Read more

Hey everyone 👋

If you’ve been anywhere near tech or marketing over the past few months, you’ve probably heard the term GEO thrown around nonstop.

• “LLM SEO”

• “Get your startup into ChatGPT”

• “Optimize for AI answers”

And now you’ve likely seen Sam Altman’s latest tweet: OpenAI is starting to test ads inside ChatGPT.

At first glance, this feels like just another monetization update. Think again maybe.

This is the moment GEO stops being a niche growth hack and starts becoming a core distribution channel. Real incentives, real budgets, and real winners. Same (or more) than what SEO has been the past 10 years.

Today’s post is not about:

• how to “rank” in ChatGPT,

• or quick tricks to get mentioned once.

Instead, we’re going to cover three things that actually matter now:

1. What GEO is and how it really works under the hood (and why most advice online is wrong)

2. Why ads accelerate GEO instead of killing it

3. What you can do today to win AI-native visibility over the next 12–24 months

If you’re building or marketing a product, this shift matters more than any Google update in the last decade, trust me.

Let’s break it down.

Join OpenAI for Startups

Sponsor Product Market Fit & get your product in front of thousands of founders worldwide? Reach out to g@guillermoflor.com

Trending right now: The Data Room Template VCs love, 27 Most Promising AI Startup Pitch Decks Backed by Top Investors in 2025

Read more

Right before Christmas Dharesh Shah’s (Billionaire founder of Hubspot) gave us a big present: all his learnings about how to position your startup in LLMs.

So I decided to break it down into actionable steps every founder can use to start capitalizing the AI Search Engines.

Other resources to learn AEO: The LLM Visibility Playbook, 𝐓𝐡𝐞 𝐑𝐞𝐝𝐝𝐢𝐭 𝐋𝐋𝐌 𝐒𝐄𝐎 𝐏𝐥𝐚𝐲𝐛𝐨𝐨𝐤

What it includes:

1) The shift: from SEO pages → AI answers

2) AEO: “SEO for the AI era”

3) How to structure content so AI cites you

4) The hidden risk: low-quality AI content can hurt you

5) Your AEO operating system (what to do first)

6) Using AI agents as teammates (not magic)

7) The coming “AI App Store” for agents

8) A 14-day implementation sprint (do this exactly)

Let’s go 🔥

1) The shift: from SEO pages → AI answers

Dharmesh’s core framing is simple: you’re moving from an SEO mindset (optimize for Google’s index + humans clicking links) to an LLM mindset (optimize for systems that consume content, synthesize, and cite).

Two consequences he highlights repeatedly:

• Organic traffic is dropping in many categories because people ask ChatGPT/Perplexity/Claude instead of clicking “10 blue links.”

• AI answers create a binary outcome: you’re either in the answer (and get disproportionate visibility) or you’re basically invisible.

• Conversion rates appear to be higher in AI searches

2) AEO: “SEO for the AI era”

AEO (Answer Engine Optimization) = optimizing your content so AI answer engines can find it, extract it, and cite it in responses.

Dharmesh’s “first-principles” rule didn’t change:

• In SEO: “The best way to rank is to be rank-worthy.”

• In AEO: be cite-worthy: don’t try to game the model; earn the citation by being the best source.

3) How to structure content so AI cites you

Read more

Hey everybody, welcome back to PMF 🔥

Subscribe now

Join OpenAI for Startups

Sponsor Product Market Fit & get your product in front of thousands of founders worldwide? Reach out to g@guillermoflor.com

Trending right now: The Data Room Template VCs love, 27 Most Promising AI Startup Pitch Decks Backed by Top Investors in 2025

While yesterday Lovable took over the timeline, this week Glean also achieved a massive acomplishment: they reached $200M ARR in only 9 months since reaching $100M ARR.

What’s super interesting about this, beyond the obvious massive and fast revenue growth, is that they are one of the few companies that are being successful with enterprise AI.

Most enterprise GenAI efforts still don’t move the needle. A recent MIT-backed report found ~95% of GenAI initiatives show no measurable impact on business outcomes. The problem isn’t the models. It’s integration, governance, and workflow adoption

Now, if you’re reading this you must be thinking: what’s in it for me?

Well, Glean achieved $200M ARR selling enterprise AI with to more than 5 different industries and 6 different use cases.

So, I decided to break Glean down and give so you can choose one of these and build a +$10M ARR Enterprise AI.

What it includes:

1. The AI Retail Playbook

2. The AI Higher Industry Playbook

3. The AI Healthcare Playbook

4. The AI Financial Services Playbook

The Enterprise AI Blueprint: 4 +$50M ARR Businesses You Can Copy From Glean

Subscribe now

The problems:

Retail looks simple from the outside: sell products.

Inside, it’s chaos: omnichannel customers, fragmented systems, high turnover, and a constant need to move fast. Glean’s own framing for retail is basically: connect the tools, make the knowledge usable, and remove the “tool switching tax.”

So retail’s core problem is not “AI content”.

It’s knowledge fragmentation at massive scale.

The exact types of retail companies to target:

Pick one, because each has different systems and pain:

1. Omnichannel specialty retail (app + stores): fashion, electronics, cosmetics, sports retail (Zara, H&M, Uniqlo, Mango, Best Buy, MediaMarkt, Fnack, Currys, Sephora, Decathlon, Foot Locker)

2. Big box / grocery: huge supply chain + store operations complexity (Waltmart, Costco, Target, Tesco, Carrefour)

3. DTC brands that became omnichannel: started Shopify, now have stores and call center

4. Marketplaces: more complexity in catalog, seller operations, fraud (Amazon, eBay, Etsy, Rakuteen, Mercadolibre)

Who feels the pain daily

• Customer support teams

• Store associates

• Store managers

• Marketing and advertising teams

• Merchandising teams

• Supply chain and logistics teams

• Engineering/product teams (context-switching hell)

The use cases:

1) Transform customer service experience

e

Problem: customers want instant, consistent answers (returns, shipping, warranty, order changes). Agents waste time searching policies and order context.

What the AI does

• Gives step-by-step resolution paths (refund vs exchange vs store credit)

• Pulls order + customer context

• Drafts the response

• Suggests upsell (accessories, bundles) when relevant

Users

• Customer support reps

• Store associates helping customers in store

• Store managers approving exceptions

2) Deliver on omnichannel

Problem: “omnichannel” isn’t a strategy anymore. It’s table stakes. Customers buy across store, web, app, social, then complain via support.

What the AI does

• Builds a customer 360 view (purchase history, browsing, loyalty, support tickets)

• Keeps product info consistent across channels

• Helps staff answer: “Do we have it in stock?” “Can I return online in store?” “Where is my order?”

Users

• Store associates + customer service reps

• Marketing teams (campaign decisions using cross-channel insights)

3) Transform employee experience (onboarding + training)

Problem: seasonal hiring + turnover. New associates are slow, mistakes are expensive, training doesn’t scale.

What the AI does

• “Ask anything” onboarding inside chat

• Instant product knowledge and process lookup

• Self-serve HR/compliance questions (vacation, safety policies)

Users

• Seasonal employees

• HR business partners

• Store managers

4) Innovate faster

Problem: retail needs fast experimentation (pricing, promotions, assortment, new channels). Teams spend forever hunting context.

What the AI does

• Merch teams run agents that summarize trends and consumer shifts

• Engineers find prior decisions, docs, tickets, code context faster

• Faster project ramp-up and fewer repeated mistakes

Users

• Merchandising team

• Engineering/product teams

5) Optimize operational efficiency and visibility

Problem: inventory, pricing, supplier contracts, logistics are complex, and legacy systems slow everything down.

What the AI does

• Summarizes inventory + pricing data for faster decisions

• Pulls insights from supplier contracts and reports

• Helps teams search across structured + unstructured ops data (emails, warehouse records, reports)

Users

• Retail ops staff

• Supply chain and logistics teams

The problem

Higher education is basically a service business with:

• A complex “customer journey” (prospect → applicant → enrolled → graduated → alumni)

• A ton of rules (financial aid, visa, grading policies, accreditation, privacy)

• Lots of stakeholders (students, faculty, staff, parents)

• Old systems (SIS, LMS, HR, finance, helpdesk)

The daily pain is:

• Students asking the same questions 1,000 times

• Staff hunting for the “latest policy” in PDFs, drives, emails

• Admin processes that require strict accuracy

The exact types of higher-ed orgs to target

1. Large universities (many departments, huge support volume)

2. Community colleges (high support load, lean staff)

3. Online-first schools (support + onboarding is everything)

4. Private universities (care about student experience and retention metrics)

Who feels the pain daily

• Admissions + enrollment

• Financial aid office

• Registrar

• Student services + advising

• IT helpdesk

• Faculty support

• HR + compliance teams

The use cases

1) Student support deflection (the big one)

What students ask all day

• “How do I register?”

• “When is the deadline?”

• “What forms do I need?”

• “How do I change my schedule?”

• “What’s the financial aid status?”

Read more

This interview is not speculative futurism, but a single argument, stated repeatedly from different angles, all around the same theme:

Human labor is the core bottleneck our civilisation has been built around, and that bottleneck is now being removed.

We argue about which jobs go, which jobs stay, how to retrain people, which skills matter… But what if human labor wasn’t the organising principle of the system?

That’s the lens through which Elon’s interview with Nikhil actually makes sense. On the surface he jumps from X to money to energy to AI to the meaning of life. Underneath, he’s doing something much more consistent:

• money stops making sense in the way we use it

• deflation is not a risk but a consequence

• energy replaces capital as the binding constraint

• coordination becomes more valuable than execution

• truth stops being ethical and becomes infrastructural

This memo follows that exact chain, without deviation.

1. Money is just a routing layer for labor

If you missed this part of the conversation, keep reading

Read more